Relational and Overt Aggression in Middle Childhood: A Comparison of Hypothetical and Reported Conflicts

Transcription

1 Illinois Wesleyan University Digital IWU Honors Projects Psychology 2001 Relational and Overt Aggression in Middle Childhood: A Comparison of Hypothetical and Reported Conflicts Jillian M. Denoma '01 Illinois Wesleyan University Recommended Citation Denoma '01, Jillian M., "Relational and Overt Aggression in Middle Childhood: A Comparison of Hypothetical and Reported Conflicts" (2001). Honors Projects. Paper This Article is brought to you for free and open access by The Ames Library, the Andrew W. Mellon Center for Curricular and Faculty Development, the Office of the Provost and the Office of the President. It has been accepted for inclusion in Digital IWU by the faculty at Illinois Wesleyan University. For more information, please contact digitalcommons@iwu.edu. Copyright is owned by the author of this document.

2 Conflict and Aggression 1 Running head: AGGRESSION IN TWO TYPES OF CONFLICT Relational and Overt Aggression in Middle Childhood: A Comparison of Hypothetical and Reported Conflicts Jill M. Denoma Illinois Wesleyan University

3 Conflict and Aggression 2 Abstract Following recent research patterns in childhood conflict, the current study examined individual differences and gender trends in conflict resolution styles. Relational and overt aggression were investigated in 31 fourth and fifth graders by use of a multi-method evaluation that included peer and teacher ratings, and hypothetical conflict vignettes and reported conflicts. It was hypothesized that girls would use relational aggression more often than boys and that boys would display overt aggression more often than girls. Teacher and peer measures were convergent in corresponding ratings of overt aggression, but no convergence was apparent for either overt or relational aggression between hypothetical and reported conflicts of both aggressive resolution strategies. Gender differences in relational aggression emerged in reported conflicts. Large effect sizes were computed for many of the tests of gender differences (overt: peer ratings, teacher ratings, reported conflicts; relational: reported conflicts).

4 .. Conflict and Aggression 3 Relational and Overt Aggression in Middle Childhood: A Comparison of Hypothetical and Reported Conflicts Conflict exists within all human relationships and is a central force in developmental change (C. U. Shantz & Hartup, 1992) because it requires individuals to use complex social skills to integrate personal desires and the wishes of others. Although conflicts arise frequently without serious consequences for relationships, they can present recurrent problems for individuals who do not possess strong conflict resolution skills. Some children use aggression as a means to solve conflict. Unlike many other approaches used to resolve normative conflict, however, the use of aggression by children has been linked to many negative developmental outcomes such as peer rejection (Dodge, et. al, 1990) and serious maladjustment later in life (C. U. Shantz, 1986). In order to prevent these outcomes from occurring, it is important to understand exactly how and why certain children use aggressive strategies to resolve conflicts while others do not. If researchers are able to more fully determine the causes and consequences of the use of aggression in conflicts, they may be able to design and implement effective intervention programs aimed at reducing the negative effects of aggression. Researchers such as C. U. Shantz (1987) and Hartup (1992) have devoted much of their professional life to studying

5 Conflict and Aggression 4 interpersonal conflict in childhood and adolescence, and have identified common causes, resolution strategies, and consequences of these conflicts. Other researchers, like Crick (2000), have focused on the construct of aggression, and its gender-specific correlates. Although previous research has investigated many issues of interpersonal conflict (i.e., precipitators, duration, frequency, outcomes) and aggression (primarily its causes, manifestations, and consequences) in children, it has not thoroughly addressed other important aspects of the topics (such as trends in age and gender behaviors and hypothetical versus reported conflict resolution styles). The current study investigates the use of aggression employed during childrens' conflicts. It expands on previous research regarding conflict resolution by investigating trends in middle childhood, a group that has not yet been effectively assessed. In an effort to thoroughly appraise types of aggression and their use in conflict resolution, the current study evaluates these concepts using a multi-method approach, which includes the use of questionnaires, hypothetical vignettes, and verbal reports of specific, real-life conflicts. It was predicted that aggression would not be used to solve most conflicts, and that boys and girls would use it almost equally. It was predicted, however, that boys and girls would typically

6 Conflict and Aggression 5 use different forms of aggression, namely overt and relational, respectively. Finally, it was hypothesized that children would report the use of similar conflict resolution strategies in hypothetical and reported conflicts, and that individual differences in aggression would be reflected across the four measures (hypothetical conflict vignettes, reports of actual conflict, peer ratings, and teacher ratings). The following sections provide a review of the current, relevant research in the fields of interpersonal conflict and aggression in childhood. After reviewing the literature presently available on these topics, the current study will identify some of the gaps in past research, and attempt to expand existing knowledge in gender trends of aggression and conflict resolution. Interpersonal Conflict Interpersonal conflict has been described as a state of resistance or opposition between individuals (Hartup & C. u. Shantz, 1992). Although mutual opposition (incompatibility) is generally accepted by scholars as the basis of conflict, debate exists over the specific structure and process involved in this type of disagreement. Some researchers, such as Garvey (1984) view conflict as a unilateral, two-part event in which person A's behavior provokes an objection or refusal from person B (as cited in Laursen & Collins, 1994). An example of this

7 Conflict and Aggression 6 definition of conflict would be as follows: John tries to take Mike's toy away from him, but Mike puts up a fight to prevent John from doing so. Instead, other researchers argue that conflict is a three-part event that requires a dyadic state of mutual opposition in which person A's behavior provokes an objection from person B, and then person A responds by persisting in the original, countering behavior (C. U. Shantz, 1987). This definition could be exemplified by the following scenario: John tries to take Mike's toy away from him, Mike puts up a fight to prevent John from doing so, but John continues to attempt the change of possession. Regardless of the number of steps defined in a conflict occurrence, researchers agree that the most critical feature in identifying interpersonal conflict behavior is the presence of individuals that possess genuinely incompatible goals. For purposes of this study, conflict is described as an interpersonal, two-part event in which person A's behavior first provokes an objection or refusal from person B. It is further be described as an incident in which the participants have genuinely incompatible intentions. This operational definition allows for reliable conflict detection by researchers and participants, as well as separation of conflict from other associated constructs like aggression and rough-and-tumble play, each of which are related to, but include unique distinctions

8 Conflict and Aggression 7 from, interpersonal conflict. Precipitators of Conflict Since conflict can be defined as one person's obstruction of another person's goal or desire, it is important to examine the types of events that most frequently provoke conflict. In young children, the majority of conflicts involve disputes over either the possession or usage of objects (C. U. Shantz, 1987), or over the control of a person's behavior (Hartup, et al., 1988). This latter type of conflict may include one person's attempt to control another's beliefs, ideas, and actions. Some researchers have reported that as children grow older, fewer of their conflicts involve objects and space, and more of their disputes focus on the control of the social environment (C. U. Shantz & D. W. Shantz, 1985). Although the previously mentioned circumstances have been found to most frequently cause conflict in children, a variety of other situations can also provoke disagreement. Most children are able to identify the primary cause of their conflicts. In a study by C. U. Shantz (1993), ninety-six percent of second grade children voluntarily stated the issue at hand when asked to describe a recent conflict. However, the event that initiates a conflict may not always continue to be the issue of contention throughout the course of the argument (C. U. Shantz, 1987). For example, if a conflict originates

9 Conflict and Aggression 8 because two children disagree about who can control the television remote, the preliminary issue of dispute (the control of the television) may be put aside during the conflict and replaced by a new disputable topic (such as who has better taste in television shows). Duration and Frequency of Conflict The majority of conflicts are brief. The average duration of conflicts across several studies of preschool aged children was twenty-four seconds (as cited in C. U. Shantz, 1987). Although conflicts are typically short, they sometimes include brief interludes, after which they occasionally resume. When investigating verbal conflicts, Eisenberg and Garvey (1981) reported that ninety-two percent of preschoolers' disputes include less than ten exchanges between partners, and sixty-six percent include fewer than five. Since the varieties of conflict are infinite, it is difficult to make accurate generalizations about the specific structure of individual conflicts. However, Hay (1984) found that preschool children are involved in approximately five to eight conflicts per hour. Laursen (1995) stated that an average of only seven conflicts per day was reported by adolescents, who more selectively chose oppositions and relationships than younger children do. The frequency of conflicts during middle childhood has not been previously documented.

10 Conflict and Aggression 9 Resolution of Conflict Certain patterns of childhood conflict resolution have become apparent in previous research studies. Laursen and Collins (1994) reported that most conflicts resulted in the definition of a distinct winner or loser, and included the disengagement or submission of at least one party. According to C. U. Shantz (1987), however, most conflicts among children are solved quickly, with relatively few instances of adult intervention. Additionally, she found that conflicts usually end with one of the following three outcomes: the clear emergence of a winner or loser, partial equality of conflicting parties (where one party concedes more than the other), or complete equality of conflicting parties (where an equal compromise is reached). Although conflicts can be resolved in a variety of ways, certain resolution styles frequently emerge. Researchers have used somewhat different names to describe these categories but there is considerable commonality across systems. Chung and Asher (1996), for example, identified and examined five conflict resolution strategies in children: adult-seeking, passive, assertive, prosocial, and hostile/overtly aggressive. Because the current research is examining an additional construct, relational aggression, six categories of conflict resolution styles will be appraised in this study: adult-seeking, passive,

11 Conflict and Aggression 10 assertive, prosocial, hostile/overtly aggressive, and relationally aggressive. Adult-seeking behaviors are those in which children tend to appeal to parents, teachers, or other people in authoritative positions when trying to resolve a conflict. Passive techniques include retreating and/or quietly sacrificing one's desires to avoid confrontational episodes. Assertive methods are those that involve a child who states his/her opinions and rights with clear and direct language during a conflict situation. Prosocial behaviors include attempts to compromise during conflict, to gratify both parties' desires, and to take both participants' feelings into consideration. Hostile techniques include many forms of overt aggression such as grabbing, hitting, and punching, verbal domination, and threats of these behaviors. Finally, relationally aggressive strategies are those in which actions that threaten to negatively affect relationships (such as implementing social exclusion, starting rumors, and denying certain friendships) are present. For purposes of this study, it was necessary to modify the list of conflict resolution styles used by Rose and Asher (1999) in order to account for the newly identified category of relational aggression. As detailed below, Crick and her colleagues (Crick, 2000; Crick & Grotpeter, 1995) have recently made a distinction between overt and covert forms of aggression.

12 Conflict and Aggression 11 To accurately assess all forms of conflict resolution, the inclusion of this new category is essential. Hypothetical Versus Reported Conflict Questions have been raised regarding the extent to which the resolution styles preferred in hypothetical conflicts are similar to those actually used in real life conflicts. Because hypothetical conflict assessment is frequently used, it is important to verify the validity of this assessment technique. In response to this debate, Sternberg and Dobson (1987) stated that the patterns of conflict resolution styles in college students for reported conflicts were similar to those found in previous research for hypothetical conflicts. They also argued that individuals' patterns of resolution for reported and hypothetical conflicts were consistent. Reinisch and Sanders (1986) also reported evidence that questionnaires regarding hypothetical conflict situations positively correspond with frequency of aggressive acts in adolescence. Chung and Asher (1996) and Dodge and Frame (1982) have claimed that available research suggests that fourth and fifth grade children's responses to hypothetical situations correspond with their actual observed behavior (as cited in Rose & Asher, 1999). Not all researchers, however, have found consistency between reported and hypothetical conflicts. For example, Iskandar, et al. (1995) reported that preschool aged children

13 Conflict and Aggression 12 more often cite the use of interpersonal negotiation (a process that is considered similar to the prosocial category used in this study) during interviews that follow hypothetical situations than they were actually to use negotiation in reallife conflicts. Additionally, Laursen, et al. (1996) reported that while hypothetical conflicts tend to emphasize mitigation, actual disagreements are resolved more often by coercion than by compromise. It is possible that the variation between reported conflict resolution styles and those observed to be employed during actual conflict behavior may be partially attributed to a social desirability bias (this bias may be reflected in the increased likelihood of reporting compromise versus actually using it). Although most previous research has suggested that resolution styles used in hypothetical conflict situations are similar to those used in real-life and reported conflicts, this assumption has never been directly assessed using hypothetical vignettes and real-life conflict reports simultaneously. For example, Sternberg and Dobson (1987) report that their results regarding conflict resolution styles used in reported conflicts correlate with those that had been reported by researchers who used hypothetical scenarios in previous studies. The current study will investigate this claim, currently based on correlational results reported by different researchers, by

14 11 Conflict and Aggression 13 simultaneously comparing the children's responses to both hypothetical and reported personal conflict. Aggression Over the years, the study of aggression has been one of the most popular areas of study in psychology. Aggressive behaviors are generally described as those that are intended to hurt or harm others (Berkowitz, 1993; Crick & Grotpeter, 1995). Since aggression has been linked to many negative developmental episodes, as mentioned earlier, researchers have argued that exploring and understanding aggression is a necessary step in preventing these potentially negative consequences from occurring. Conflict and aggression are linked in ways that are often misunderstood. Although aggressive displays often occur during conflicts, the majority of conflicts do not involve aggression (C. U. Shantz & Hartup, 1992; C. U. Shantz, 1987). Furthermore, aggression can be distinguished from conflict in that aggression is only one of many ways to solve a conflict situation. Research has shown that certain conflict situations, including disagreements over objects, threats to one's ego (Hartup, 1974), and social manipulations (Perry et al., 1992), may be more likely than others to provoke aggression. Almost all of the traditional aggression research concludes that boys are more aggressive than girls (Crick, 2000;

15 Conflict and Aggression 14 Berkowitz, 1993; Miller, Danaher, & Forbes, 1986; Reinisch & Sanders, 1986; Parke & Slaby, 1983). Because of this assumption, the majority of studies on aggression have occurred with male subjects, and have primarily assessed overt forms of aggression. Recent studies, however, have also begun to focus on a newly identified form of aggression called relational aggression that has been reported to be more prevalent in girls (Crick, 2000; Crick & Grotpeter, 1995). The distinction between these two forms of aggression, overt and relational, is critical to understanding the frequency of aggressive acts in children. For example, Crick, et al. (1999) concludes that gender differences in aggression are minimal when both overt and relational forms of aggression are considered. It is also important to make the distinction between these two forms of aggression when attempting to identify the separate consequences and intervention programs involved with relational and overt aggression, respectively. Overt Aggression A principle reason why researchers have focused on physical aggression is that overt behaviors are much more easily identified and assessed than are covert behaviors. Overtly aggressive behaviors are those that harm others through physical damage or through the threat of such damage (Crick, 2000; Crick

16 Conflict and Aggression 15 & Grotpeter, 1995). Acts like kicking, punching, hitting, loud yelling, and threats of violence are more apparent, and therefore, more likely to be noted by observers, than are the camouflaged indirect, interpersonal, or covert behaviors that fall in the relationally aggressive category. Certain types of conflict situations have been identified as being more likely to evoke overtly aggressive responses than others are. Results of previous studies have suggested that instrumental conflict situations (e.g., having a science project purposefully destroyed by a peer or being cut in front of in line by another peer) most often elicit overtly aggressive responses than other types of conflict (as cited in Crick & Werner, 1998). Relational Aggression Although most aggression research has focused on identifying and examining the incidence of overt behavior, many recent studies have begun to also explore relational aggression. Several decades before this contemporary research was done, Feshbach (1969) identified a similar construct she labeled "indirect aggression," which included behaviors such as rejection and social exclusion. Contemporary researchers distinguish between indirect and relational aggression in saying that indirect aggression does not include all forms of relationship manipulation (both indirect and direct forms),

17 Conflict and Aggression 16 whereas relational aggression does (Crick, et al., 1999). Currently, researchers describe relationally aggressive behaviors as those which specifically serve to harm relationships (Crick, 2000). This form of aggression can manifest itself in many forms including peer manipulation, rejection, character defamation, and social exclusion. According to several studies, girls exhibit significantly higher levels of this form of aggression than boys (Crick, 2000; Cairns, Cairns, Neckerman, Ferguson, & Gariepy, 1989; Feshbach, 1969), and they compose the majority of groups defined as relationally aggressive (Rys & Bear, 1997). Additionally, girls look more favorably upon using relationally aggressive techniques in solving relational conflicts than boys (Crick & Werner, 1998). Why are girls more likely to exhibit relationally aggressive behaviors than boys? Although the answer to this question has not yet been thoroughly examined, some initial hypotheses have been advanced. One line of reasoning takes into account the difference in physical structure between boys and girls. Lagerspetz, et al. (1988), proposed that since males usually possess a bigger stature and size than females, they can generally afford to be more physically confrontational. He further stated that girls realize they may not be able to actively defend themselves in direct conflict, so they have

18 Conflict and Aggression 17 learned ways of harming others that do not risk direct retaliation. During middle childhood, however, it is not rare for girls possess physical stature similar to that of boys. With this fact in mind, the gender-based rationale presented by Lagerspetz, et al. (1988) may not be applicable to elementary school-aged children. Another body of research suggests that since aggression is defined as any behavior that is intended to hurt others, and since interpersonal relationships are generally more important to girls than to boys, it logically follows that one of the most effective ways to hurt a girl is to cause damage to an interpersonal relationship of hers (Crick & Grotpeter, 1995). Several negative outcomes have been identified in children who frequently use relational aggression. Crick and Grotpeter (1995), reported that relational aggression was significantly related to social maladjustment (e.g., peer nominations of rejection and self-reports of poor peer acceptance), loneliness, depression, and social isolation. This finding suggests an urgent need for a greater understanding of relational aggression. Researchers must attempt to develop intervention programs that specifically address the potential socialpsychological maladjustment that may be present in relationally aggressive children. As noted earlier, situations such as disputes over

19 Conflict and Aggression 18 instrumental conflict topics, were more likely than others to promote overtly aggressive responses in children who frequently use aggression as a means to resolve conflicts. Similarly, it has been reported that relationally aggressive children tend to attribute hostile intent to peers in contexts that include ambiguous, negative relational events (e.g., not receiving an invitation to a friend's birthday party or discovering that a friend is playing with a disliked peer) (Crick & Werner, 1998). Although recent literature is beginning to more clearly define the actual construct of relational aggression, little research has attempted to investigate the role of relational aggression in conflict resolution. Current Study The current study approaches the assessment of individual differences in relational aggression levels and conflict resolution styles using a multi-method evaluation. General measures of aggression are evaluated using a peer rating scale and a teacher rating scale. Overt and relationally aggressive conflict resolution strategies are assessed by two self-report measures: hypothetical vignettes and reported conflicts. These instruments are used in an effort to more fully understand the construct of relational aggression, its rate of frequency in conflicts, and the types of subjects and conflict situations with which it is most likely associated.

20 Conflict and Aggression 19 This research assesses the correlation between overt and relational aggression on each of the four measures. It was expected that high scores on global measures of overt and relational aggression (peer and teacher ratings) would correspond with high scores of overt and relational aggression on the self-report measures of conflict resolution style (hypothetical vignettes and reported conflicts), respectively. It was predicted that teacher and peer ratings of both overt and relational aggression would be comparable. It was further expected that, across the two measures of self-report, there would be convergence of both the overt and relationally aggressive conflict resolution strategies. There were many hypothesized gender differences for this study. It was predicted that girls and boys would exhibit similar overall ratings of aggression on teacher and peer ratings, since both relational and overt forms were be studied. It was predicted, however, that when relational aggression was used by an individual as a means to resolve conflict, it would more often be used by girls than by boys. In contrast, it was hypothesized that boys would use overtly aggressive conflict resolution strategies more often than girls. Method Participants The subjects consisted of 31 fourth and fifth grade

21 Conflict and Aggression 20 students (20 females, 11 males) from a suburban Midwestern U.S. elementary school. There were 18 fourth grade students and 13 fifth grade students. Measures Aggression Peer Ratings. A portion of the peer nomination instrument developed by Crick (1995) and Crick and Grotpeter (1995) was used to assess subjects' relational and overt aggression levels. The inventory consisted of five relational items (e.g., kids who try to keep certain people from being in their group when it's time to play) and five overt items (e.g., kids who shove and push others around). For each item, the subject was asked to rate every participating classmate according to how strongly he/she fit the description. Additionally, nine filler items were added (e.g., kids who are smarter than most). The aggression scores were computed by summing and standardizing the scores each child received on the relational and overt scales, respectively (see Appendix A for complete measure). Each participant was rated by three to five children, depending upon the number of participating classmates. Teacher Ratings. Teachers rated overt and relational aggression on Crick's inventory rating the extent to which each participating child exemplified the behavior described each item. This measure was identical to the peer rating, except

22 Conflict and Aggression 21 that it did not include any filler items. This assessment was compared to the peer appraisal of each child's aggression level to determine whether the results from the two different groups were positively correlated. As Crick reported (2000), these scales have been found to possess high internal consistency (Cronbach's Alpha ranged from.82 to.89 for relational aggression and from.94 to.97 for overt aggression). Additionally, the test-retest reliability score for a four-week interval was.82 for the relational scale and.90 for the overt scale. In a factor analysis, two distinct factors emerged (relational and overt) with scores ranging from.73 to.91. Furthermore, the cross-loadings were moderate (r.54), which provided evidence that two separate, yet related factors were present. Conflict Resolution Style Students completed two different tasks that assessed their conflict resolution behaviors. In an effort to investigate the hypothesized difference in description between hypothetical and reported conflicts, and to most accurately assess each subject's resolution style, two different measures were given to participants. The first measure appraised the children's conflict resolution styles in hypothetical situations, and the second one assessed individuals' resolution choices in reported conflicts that had occurred in their lives.

23 Conflict and Aggression 22 Hypothetical Conflicts. To assess which style participants favor in hypothetical conflict situations, a measure based on the Children's Conflict Resolution Measure was created (Chung & Asher, 1996). The original measure consisted of twelve vignettes that depicted cornmon conflict situations of children (e.g., disputes over the use of toys and how to spend free time). In this study, children were read eight vignettes that detailed realistic, age appropriate, social conflict scenarios (see Appendix B). Several of these vignettes were ones used by the original researchers, while a few were modified slightly in order to provide for easy cross-cultural comparisons that will be made a later date. After ensuring comprehension of each vignette, the subjects were asked to imagine themselves being involved in the situation with a classmate, and then asked to rate how likely they would be to use each of six possible responses using a five point scale (each of the possible choices was listed randomly and corresponded with the six, previously identified resolution strategies - adult-seeking, passive, assertive, prosocial, hostile/coercive, and relational). This process of reading the vignettes and possible resolution strategies closely mirrored that used by Chung and Asher (1996) (see Appendix C). Scores within each category were averaged from the subject's ranking on all vignettes.

24 Conflict and Aggression 23 Although each of the resolution style responses had been adapted to the specific vignette, each retained a consistent structure that corresponded closely with the description of each category. Adult-seeking responses consisted of subject's appeal to an authority figure when resolving conflict. Passive responses involved a retreat, and surrendering of one's wants. Assertive methods involved the subject's direct statement of personal rights or desires. Prosocial responses incorporated the needs of both conflict participants and ended in some form of a compromise. Hostile/coercive responses involved overt aggression of either the physical (hitting, shoving, grabbing) or verbal type (yelling). Finally, relational responses involved the threat or action of destroying another's relationships, social standing, or reputation. For specific examples of responses that fall into each of these categories, see Appendix D. Although the modified measure used in this study had not been formally evaluated for psychometric properties, Chung and Asher (1996) reported good psychometric characteristics for a similar procedure. Internal consistency assessment was highest for prosocial and hostile/coercive strategies (Cronbach's alpha coefficients were.79 and.90, respectively). However, the internal consistency for the adult-seeking, passive, and assertive categories was somewhat lower (providing coefficients

25 Conflict and Aggression 24 of.57,.55, and.40 respectively). No data was available on the relational aggression response component. Reported Conflicts. For the second conflict resolution style assessment, children were asked to report recent conflicts that had occurred in their lives. On each of three occasions, participants were asked to recall an interpersonal conflict that had taken place within the last 3 days and to describe it in detail. The participants were specifically asked to think of a recent conflict in their own life, and to visualize this event as it occurred. From start to finish, children were asked to recall the event, step-by-step, as it happened. The researchers were able to ask several prompting questions regarding who the conflict involved and how it was resolved (see Appendix E) to clarify the situation's details and to compile all of the desired information, if necessary. This procedure was modeled after the one described by C. U. Shantz (1993). The first reported conflict assessment occurred in person during the initial interview appointment, to assure that all participants fully understood the directions of this task. On the subsequent two occasions, participants were asked to report a recent conflict over the phone. It was beneficial to ask children to mention recent conflict episodes because doing so may have helped the participant to improve the accuracy of their memory and to minimize recall error.

26 Conflict and Aggression 25 Before completing each of the three interviews, children were reminded of the operational definition of conflict: a serious, interpersonal, two-step event where person A's behavior provokes an objection or refusal from person B. This operational definition was described to each participant in appropriate terms to ensure their understanding of it. Children were told that an interpersonal conflict occurs when ~two or more people have different ideas, opinions, or wishes. These differences may cause the people to argue, disagree, or fight with one another. H They were also given examples of several types of conflict and told that conflicts could range from a simple difference in opinion to an all out physical brawl. This measure yields qualitative data that was coded independently by two researchers. First, researchers determined a primary resolution strategy. Secondly, they identified all other contributing resolution strategies apparent in the conflict. The primary style was transposed into both dichotomous and continuous values. Procedure The students were selected according to an informed, voluntary consent procedure. They were initially recruited by sending horne a letter explaining the study to the parents of all fourth and fifth graders in the targeted school. The information included the purpose of the study, the details of

27 Conflict and Aggression 26 its procedure, and discussed the potential risks and benefits of the participants. The parents who returned a prepaid postcard (31%) were contacted by telephone. If the parent granted approval of participation and scheduled an appointment, informed assent was also obtained from the participant. The interviews were conducted at the participant's school during after school hours. Following the completed consent of the parent and assent of the subject, the interviewer administered the peer rating scales and the hypothetical conflict vignettes. During this first appointment, the researcher also asked the participant to describe one recent conflict in their life. Follow-up phone calls were made to the participants at weekly intervals. This process was mirrored after the one detailed by Patterson, Reid, and Dishion (1992). They stated that phone interviews provided a stable estimate of children's behavior. On each of these subsequent occasions, the interviewer followed a standard format for phone interviews developed by Jones (1974) which called for a "matter-of-fact style with minimal personal conversation" (as cited in Patterson, et al., 1992). The researcher manually transcribed the interviews. All coding was done by two undergraduate students. They independently coded 100% of the transcripts. Any discrepancies between the two coders were verbally discussed until consensus

28 Conflict and Aggression 27 was obtained. Results High internal consistency was found for peer ratings, teacher ratings, and hypothetical conflict situations. Cronbach's alpha for overt aggression was.83,.96, and.80 for peer ratings, teacher ratings, and hypothetical conflict measures, respectively. Cronbach's alpha for relational aggression was.80,.90, and.86 as assessed by peer ratings, teacher ratings, and hypothetical measures, respectively. The following sections report correlations of overt and relational aggression between measures. All correlation coefficients can be seen on Tables 1 and 2 for overt and relational aggression, respectively. Within Measure Overt and Relational Aggression Comparisons Pearson's correlation coefficient was computed to assess the relation between overt and relational aggression within measures. Significant correlations between overt and relational aggression emerged within peer ratings, ~ =.83, E <.01, teacher ratings, r =.75, E <.01, and the hypothetical vignette instrument, ~ =.84, E <.01. Teacher and Peer Ratings of Global Aggression Overt and relational aggression as assessed by peer ratings and teacher ratings were analyzed using Pearson correlations. A significant positive correlation emerged between peer and

29 Conflict and Aggression 28 teacher measures of overt aggression, ~ =.53, 2 <.05; however, the correlation between these ratings of relational aggression was not significant, ~ =.17, ns. Comparison of Conflict Resolution Between Vignettes and Reported Conflicts Overt and relational aggressive conflict strategies obtained from the assessment of hypothetical and reported conflicts were compared. Each participant's overt and relationally aggressive resolution scores from the reported conflicts were computed in two different ways. First, each student's score was calculated as a percentage of the times he/she reacted to a conflict using an overt or relationally aggressive response. Then a dichotomous score was also computed for each aggression category. Each child was coded as to whether or not the use of overt or relational aggression in a conflict was reported. Both Pearson's correlation (for continuous variables) and point biserial correlations (for dichotomous variables) were used to compare overt and relationally aggressive conflict resolution strategies across these measures. On the hypothetical vignette measure, an average score of 1.50 (SO =.50) was received on the overt aggression category and an average score of 1.59 (SO =.64) was received on the relational aggression category. Of the 77 reported conflicts

30 Conflict and Aggression 29 analyzed for this study, 16 (21%) were coded as being primary overtly aggressive and 9 (12%) were coded as being primarily relationally aggressive. No significant effects emerged from the calculation of Pearson correlation coefficients comparing overt and relational aggression between hypothetical and reported conflict measures. Results within the overt aggression resolution category yielded r =.01, and results within the relational aggression resolution category yielded =.11. Likewise, the results of the point biserial method indicated a positive, but insignificant correlation, rb =.19, ns and rb =.24, ns for overt and relationally aggressive conflict resolution styles, respectively. Comparison of Global Ratings of Aggression with Conflict Resolution Strategies Peer Ratings and Hypothetical Conflict Measures. Assessments between peer-evaluated aggressive behavior and hypothetical conflict aggressive responses yielded no significant correlations. Results in the overt and relational aggression categories yielded r =.09 and r = -.06, respectively. Peer Ratings and Reported Conflict Measures. Comparisons of peer ratings and reported conflict measures yielded no significant correlations. Obtained values were r

31 Conflict and Aggression for overt aggression and r -.11 for relational aggressions. Teacher Ratings and Hypothetical Conflict Measures. No significant correlation emerged between teacher ratings of global aggression and the hypothetical conflict resolution styles of aggression. Overt and relational aggression both produced small, positive correlations, r =.24, ns and r =.36, ns respectively. Teacher Ratings and Reported Conflict Measures. No significant correlations were found between global assessments of aggression and the aggressive conflict resolution styles in the teacher ratings and the reported conflict measures, respectively. Pearson's correlation coefficient was -.04 for overt aggression and -.35 for relational aggression. Gender Differences Differences in responses between boys and girls were analyzed using t-tests. If Levene's test showed that the variance of the two groups was significantly different, the Welch-Aspen adj ustment was used. This adjustment uses a nonpooled error term and calculated degrees of freedom. In the absence of a significant Levene's Test, equal variance was assumed. For additional clarification of gender differences among the measures, also see Table 1. Peer Ratings. The mean score for global overt aggression

32 Conflict and Aggression 31 on this measure was 1.65 (SO. 75). The average score for girls (~ = 1.53, SO =.43) was lower than for boys (~ = 1.29, SO.35). A t-test produced results that were not significant,!(25) 1.23, ns. The mean score for global relational aggression on peer ratings was 1.92 (SO =.82). The mean score for girls (M = 1.91, SO =.49) was higher than for boys (M = 1. 78, SO. 40). This comparison of gender means was not significant, t(25)= -.71, ns. Cohen's d' was used to examine the effect size present for the different genders in the overt and relational aggression categories of the peer ratings. These analyses yielded a moderate result of d' =.49 for overt aggression and a small result of d' =.29 for relational aggression. Teacher Ratings. The global overt category yielded a mean score of 2.12 (SO 1.21) on the teacher rating. The average score for boys (~ 2.50, SO = 1.23) was higher than for girls (~ = 1.93, SO = 1.20), but this difference was not significant, t(16) =.94, ns. The global relational aggression score on the teacher measure was 2.58 (SO = 1.82). This score did not significantly differ between boys (M 2.57, SO 1.11) and girls (M = 2.58, SO = 1.12), t(16) -.03, ns.

33 Conflict and Aggression 32 Effect sizes were calculated by using Cohen's d'. A moderate effect size of.47 was found for overt aggression, and a small effect size of.02 emerged for relational aggression.

34 Conflict and Aggression 33 Hypothetical Vignettes. No significant gender differences were found in the overt and relationally aggressive resolution categories in the hypothetical vignettes. For overt aggression, no significant differences were found, t(29) = -.06, ns, between boys (~= 1.49, SO =.44) and girls (~ = 1.50, SO =.55). For relational aggression, there were no significant sex differences!(29) =.09, ns, and boy's means (M = 1.60, SO =.60) were slightly higher than girls (~= 1.58, SO =.66). Cohen's d' produced an effect size of.02 and.03 for overt and relational aggression, respectively. Reported Conflicts. Eleven of the 16 overtly aggressive conflicts were reported by boys and 5 were reported by girls. All 9 of the relationally aggressive conflicts were reported by girls. The Welch-Aspen test indicated that raw overt scores resulted in a insignificant difference between boys (~=.35, SO =.37) and girls (~=.13, SO =.18), t(12.70) = 2.11, ns. The effect size of this contrast was.76. Differences between the raw relational aggression scores of boys (~=.00, so =.00) and girls (~ =.19, SO =.32) were significant, t(19) = -2.67, ns, according to the Welch-Aspen test. This category's effect size was.70. Chi square tests were used to assess the potential difference between observed and expected values for overt and relational aggression as measured by dichotomous scores. Overt

35 Conflict and Aggression 34 aggression in reported conflicts was used boys 69% of the time and by girls 31% of the time. A chi square test indicated that there was not a significant gender difference for overt aggression, X 2 (1) = 2.71, ns. One hundred percent of conflicts involving relational aggression were reported by girls, and thus, a significant gender difference emerged, X 2 (1) = 4.09, ns. Discussion In this study, gender differences in aggression and conflict resolution strategies were examined using a multimethod approach. Analyses indicated that few gender differences emerged on any of the measures; relational aggression, however, was reported more often by girls than boys on the reported conflict measure. No significant correlations emerged between the two self-report measures (hypothetical vignettes and reported conflicts) for either the overt or relational aggression scale. The results in this report are the preliminary findings from the analysis of the subset of data that are currently available. At its completion, this study will include approximately 120 participants from the United States (~ = 60) and Indonesia (~ = 60). Many of the correlations and tests of gender differences done for this paper provide results that are not statistically significant at the present time. When this study is finished, however, it is possible that some of the

36 Conflict and Aggression 35 results that currently approach significance may become significant. The results from this study provide new insight into the conflicts of fourth and fifth grade students. Results are consistent with earlier claims that overt and relational aggression are related constructs. The first question of interest was the correspondence between teacher and peer ratings of overt and relational aggression. When peer and teacher ratings of overt and relational were examined, a significant pattern appeared. Both teachers and peers consistently rated the same children as either high or low in overt aggression. The current study's results also suggest that overt aggression may be more easily identified and observed by others than relational aggression. Teachers and peers were unable to agree on which children were relationally aggressive. Relational aggression is covert may not be easily seen by outsiders. Subsequent research must address the issue of more effectively identifying relationally aggressive children. Although teachers and peers converged in the identification of global overt aggression, convergence did not occur on the two self report measures of conflict resolution styles (hypothetical vignettes and reported conflicts). Children did not consistently indicate their use of overt or relationally aggressive behaviors in these two measures. The current

37 Conflict and Aggression 36 inconsistent scores of overt and relational aggression across hypothetical and reported conflicts support the claims of Laursen, et al. (1996) and Iskander, et al. (1995) that in hypothetical conflicts, children underreport hostile and manipulative resolution styles, and overreport prosocial styles. This phenomenon may be partially due to a social-desirability bias. It also is possible that the hypothetical conflict measure used in this study was not age appropriate. Ultimately, the current findings suggest that conclusions drawn about an individual's conflict resolution styles based solely upon hypothetical conflict measures should be interpreted with caution as they may not be a valid reflection of the behavior in real life contexts. It is necessary to acknowledge the potential social desirability effect that may affect children when recounting conflicts. Since the children were interacting with another individual, it is also possible that they were affected by an interviewer bias. On the other hand, since a reported conflict is a child's representation of a real-life situation, it may be that a more accurate representation of an individual's true resolution style may emerge with this type of measure. No significant correlations were found between the measures of global overt and relational aggression (teacher and peer ratings) and the measures of conflict resolution strategies

38 Conflict and Aggression 37 (hypothetical vignettes and reported conflicts). Peer ratings did not correspond well with either of the conflict resolution strategies. Although final conclusions cannot be made until the complete data set is analyzed, it may be that fourth and fifth grade children may not be able to accurately evaluate the aggressive behaviors of their peers using the rating procedure employed in this study. Part of the children's difficulties may result from the trouble they experienced with establishing norms for these behaviors. These theories may explain why Crick, et al. (1998) reports that peer-teacher rating correspondence increases with grade level. Teacher ratings also failed to correlate significantly with the measures of conflict resolution. However, there were at least modest correlations apparent between these instruments. Of particular interest is the moderate correlation of relational aggression between teacher ratings and hypothetical conflict appraisals. This correspondence will be closely examined when the complete data set is available. Although no significant correlations of overt or relational aggression were obtained many of the measures, calculated effect sizes suggest that potentially in a larger sample, certain correlations, such as teacher ratings with hypothetical conflict scores of overt aggression and teacher ratings with reported conflict scores relational aggression, may become significant.

39 Conflict and Aggression 38 It is also possible, however, that an increased sample size will not result in significant gender differences in overt or relational aggression. Despite the sample size in used in this report, some gender differences in aggressive behavior emerged. Of special interest are the results of the reported conflicts. Girls provided 100% of the reports of relational aggression. The results of the dichotomous analysis provide support for the hypothesis that girls and boys differ in their reported use of relational aggression. Although boys provided 69% of the overt aggression responses in these conflicts, the gender difference was not significant. Examining the effect sizes of these two aggression forms in reported conflicts provides some support for the assumption that with an increased sample size, significant gender differences may result. The effect sizes of peer ratings also provide evidence that a larger data sample may produce significant gender differences in aggression. The moderate effect size of overt aggression in peer ratings reported in this study, combined with the previously reported large effect sizes of overt aggression in peer ratings (Crick & Grotpeter, 1995), suggests that the complete data set for this study may reveal differences in the levels of overt aggression in boys and girls.